Pathogenetic and therapeutic perspectives on neurocognitive models in psychiatry: A synthesis of behavioral, brain imaging, and biological studies.

Neurocognitive assessments are useful to determine the locus of insult as well as functional capacities of patients on treatment. In psychiatry, neurocognitive assessment is useful in the identification of brain lesions, evaluation of cognitive deterioration over time, and advancement of theories regarding the neuroanatomical localization of symptoms. Neurocognitive models provide a bridging link between brain pathology and phenomenology. They provide a useful framework to understand the pathogenesis of psychiatric disorders, bringing together isolated findings in behavioral, neuroimaging, and other neurobiological studies. This review will discuss neurocognitive model of three disorders - schizophrenia, bipolar disorder, and obsessive compulsive disorder - by incorporating findings from neurocognitive, neuroimaging, and other biological studies.

INTRODUCTION

Neurocognitive assessments are useful to determine the locus of insult as well as functional capacities of patients in treatment. In psychiatry, neurocognitive assessment is applicable for identification of brain lesions, evaluation of cognitive deterioration over time, and advancement of theories regarding the neuroanatomical localization of symptoms.[1] Thus, neurocognitive findings enable one to develop hypotheses about functional brain abnormalities in patients with major psychiatric disorders. Neurocognitive models are based on studies examining patients with a particular disorder in comparison with healthy controls or patients with other disorders. While neurocognitive models have been proposed for different psychiatric disorders, this review will discuss three important disorders, namely, schizophrenia, bipolar disorder, and obsessive compulsive disorder (OCD).

SUMMARY OF NEUROCOGNITIVE FINDINGS IN PSYCHIATRIC DISORDERS

Obsessive compulsive disorder

Studies examining patients in symptomatic phase have reported specific executive function deficits in OCD – response inhibition, set shifting, and decision making.[2] Non-verbal memory deficits are reported in few studies, which could be secondary to deficits in organizational strategies.[3] Moreover, OCD patients possibly have a decreased degree of confidence in their memory – metamemory deficits.[4] In addition, OCD subjects have a memory bias for threatening information[5] and preliminary results suggest abnormalities in emotion processing.[6] However, studies report absence of impairment in intelligence, verbal memory, attention, verbal fluency, and planning.[2]

Very few studies have examined the neurocognitive profile of OCD patients in recovered phase. Bannon et al.[7] reported impaired set-shifting abilities in OCD patients in both symptomatic and remitted phases. To examine whether these neurocognitive deficits are state or trait markers, we examined 30 patients with OCD in recovered phase and compared with individually matched controls. Patients in recovered phase of the illness had significant deficits in tests of set-shifting ability, alternation, response inhibition, and nonverbal memory, but had intact performance in other tests. Study findings suggested trait nature of these deficits and raised the possibility of being endophenotypes[8] [Figure 1]. In another study, we examined patients in symptomatic and remitted phase using a novel optimized emotional Stroop test. This test was developed in order to control the methodological limitation of previous studies by using lexically matched words. Only symptomatic patients, but not remitted patients had higher attention bias for negative OCD stimuli, but not for neutral or non-OCD emotional stimuli. The observations were in support of threat-related hypothesis[6] [Figure 2].

Figure 1

Mean scores between recovered OCD and controls. CPT - Continuous performance task; WCST - Wisconsin card sorting task; DAT - Delayed alternation task; TOL - Tower of London; CFT - Complex figure test; AVLT - Auditory verbal learning test; Matrix - Color matrix

Figure 2

Reaction time measures of patients and controls on optimized emotional Stroop test

Studies have examined unaffected relatives of patients to know whether these deficits are shared by the family members.[910] In relation to matched healthy control subjects, relatives had deficits in cognitive flexibility and motor inhibition, but intact decision making. The profile of dysfunction on the neuropsychological tasks in unaffected relatives was indistinguishable from that of OCD patients.

Schizophrenia

Studies in schizophrenia have revealed deficits across many cognitive domains. Deficits are more prominent in working memory, episodic memory, processing speed, visuo-spatial skills, and executive functions, importantly attention, set shifting, and response inhibition.[1112] Few studies have also examined laterality abnormalities in schizophrenia and reported the absence of pseudoneglect in schizophrenia.[1314] These deficits did not have relation with symptom severity and type. Deficits were present even after the remission of clinical symptoms and even before the onset of clinical symptoms, supporting neurodevelopmental hypothesis; studies examined children and military inductees in a prospective design before the onset of illness and reported those who later developed schizophrenia had deficits in overall intelligence and also specific domains like attention and memory.[1516] In addition, family studies examining unaffected relatives of schizophrenia patients have also shown that family members share these cognitive abnormalities, but at a lesser severity. These have supported the view that neurocognitive deficits are potential endophenotypes in schizophrenia.[17]

Bipolar disorder

Studies in bipolar disorder have examined patients in manic, depressive, and euthymic phases. Patients in euthymia had deficits in executive functions predominantly involving attention, working memory, verbal memory, and executive functions (set shifting, problem solving, and response inhibition).[18-20] They also had impairments in non-verbal memory deficits and impaired visuo-spatial functions. Patients in manic phase in addition had impairments in verbal learning and memory. In depressive phase, there were impairments in attention, verbal memory, fluency, and executive functions, mainly set shifting.[18] Studies have also shown increased sensitivity to emotional stimuli and abnormal emotional perception.[2122] These deficits were seen in unaffected relatives of patients with bipolar disorder[23] and also in at-risk individuals even before the onset of symptoms.[24]

We recently examined laterality abnormalities concurrently in schizophrenia and bipolar disorder using a line bisection task. Our findings suggested attenuation of normal pseudoneglect in schizophrenia and accentuation of normal pseudoneglect in bipolar disorder, indicating lesser lateralization in schizophrenia and possibly greater lateralization in bipolar disorder[25] [Figure 3].

Figure 3

Contrasting deviation profile of right hand in schizophrenia patients, bipolar disorder patients, and controls

NEUROCOGNITION – NEUROIMAGING CORRELATIONS

In summary, results of neurocognitive studies suggest deficits in prefrontal cortex and basal ganglia. These findings are consistent with results from neuroimaging studies. Structural and functional studies have largely implicated abnormalities in prefrontal cortex and basal ganglia. There is a broad consensus for widely distributed abnormalities involving fronto-striatal circuits in OCD.[26-29] Moreover, reversal of orbitofrontal cortex (OFC) dysfunction after treatment has also been reported.[30]

A review of volumetric magnetic resonance imaging (MRI) studies concluded that volumes of regions such as the amygdale, hippocampus, frontal lobe, parahippocampus, thalamus, and superior temporal gyri are decreased in patients when compared with controls.[31] Schizophrenia-related size differences in the cerebrum have been particularly associated with the heteromodal association areas, especially the prefrontal, temporal, and inferior parietal cortex.[32] Studies examining corpus callosum (CC) in schizophrenia including a meta-analysis of 11 studies[33] reported decreased global size of CC in schizophrenia patients.

Earlier neuroimaging studies examined chronic schizophrenia patients on long-term neuroleptic treatment using low-resolution MRI which might confound results. In an attempt to avoid these potential confounds and to examine the status of brain structure in Indian schizophrenia patients, we studied antipsychotic naïve schizophrenia patients using 3 T MRI at National Institute of Mental Health and Neurosciences, Bangalore. In these studies, we examined cerebellum, CC, thalamus, and inferior parietal lobe using semi-automated morphometric techniques in comparison to matched healthy subjects. In addition, we examined the relation between clinical features and brain pathology, bridging the link between structural lesions and clinical phenotype.

In our first study, we examined cerebellar gray matter volume using voxel-based morphometry; schizophrenia patients had significant cerebellar gray matter volume loss compared to healthy controls. Importantly, cerebellar gray matter volume had correlation with negative symptoms, supporting the concept of cognitive dysmetria[34] [Figure 4]. Next, we examined CC abnormalities using gold standard manual morphometric method. Patients had significantly smaller CC, splenium, and isthmus areas than controls. A novel finding of the study was only those without first rank symptoms (FRS), but not those with FRS differed from controls. Study findings supported neurodevelopmental hypothesis and potentially different connectivity abnormalities in symptom genesis[35] [Figure 5]. In a recent study, we examined thalamus volume in relation to symptom dimensions of schizophrenia. Right, left, and total thalamus volumes of patients were significantly smaller than those of controls and volumes had differential correlation with symptom dimensions. Findings suggested contrasting pruning aberrations to underlie different symptom genesis in schizophrenia[36] [Figure 6].

Figure 4

Gray matter volume deficits in patients with schizophrenia in comparison with controls

Figure 5

Corpus callosum and its subdivisions

Figure 6

Manual tracing of the thalamus with outline in red (left) and whole volume in red (right)

Functional imaging studies have consistently demonstrated deficits in prefrontal lobe activation in schizophrenia.[3738] A functional MRI (fMRI) study has reported Schneiderian first rank symptoms to be associated with parietal lobe hyperactivity in schizophrenia patients.[39] Another fMRI study has reported temporal randomness (which is an indicator of spontaneous willed action) to correlate with frontal lobe dysfunction in schizophrenia patients.[40] Thus, we examined inferior parietal lobe volume using an interactive, three-dimensional, semi-automated analysis. Patients with first rank symptoms of schizophrenia had decreased inferior parietal lobe volume.[41] Thus, overall structural and functional imaging studies demonstrate deficits in prefrontal, temporal, and parietal lobes.

Following the seminal study by Drevets and colleagues,[42] several studies have examined and reported decreased volumes of frontal cortex in adults with bipolar disorder.[4344] Interestingly, there was also a significant association between magnitude of ventral prefrontal cortex volume decrease and rapid cycling.[44] However, studies in adolescents are inconsistent with disparate findings.[45] Studies examining the amygdale volume have been inconsistent; while some have reported decreased amygdala volume, others have reported increase or no differences.[45] However, studies in childhood bipolar disorder have consistently reported reduced amygdale volume.[46] Similarly, different studies have demonstrated decreased volumes of hippocampus,[47] cerebellum,[48] and basal ganglia.[49] Functional imaging studies have reported deficits in ventromedial prefrontal cortex (VMPFC) activation on emotion and cognitive processing tasks, with lateralization to the right in mania and to the left in depression.[45] Studies also have consistently demonstrated abnormal amygdala activation on emotional processing tasks.[50] In addition, abnormalities have been found in thalamus, striatum,[51] and cerebellum.[52]

NEUROCOGNITION – OTHER BIOLOGICAL CORRELATIONS

In a recent study, we examined ratio of length of second digit to fourth digit (2D:4D) in patients with schizophrenia and controls. 2D:4D has been put forth as a potential indicator of cerebral lateralization. Study findings revealed significantly lower 2D:4D asymmetry index in male schizophrenia patients than male controls, and those who had FRS had the lowest index. The study findings further support the cerebral lateralization theories in schizophrenia.[53] In a related study, we examined dermatoglyphic complexity index (DCI) in patients with schizophrenia and its relation with gray matter volume. Lesser DCI was associated predominantly with decreased frontal and temporal gray matter volumes. On the contrary, increased parahippocampal and inferior parietal lobule gray matter volumes (regions implicated in the genesis of first rank symptoms) were associated with lesser DCI, which might offer partial explanation to first rank symptoms being less common in younger age at onset patients.[54]

NEUROCOGNITIVE MODELS OF PSYCHIATRIC DISORDERS

Results are largely suggestive of failure in executive functions in terms of response inhibition, set shifting, and organizational strategies. Tests which measure these cognitive functions have been shown to be sensitive to dysfunction of OFC and frontal–striatal circuit.[55-57] Though lesions in dorsolateral prefrontal cortex (DLPFC) has shown perseverative errors in Wisconsin card sorting task (WCST), localization is less accurate. Deficits in non-verbal memory are possibly secondary to organizational strategies and OFC is implicated in the initialization of effective behavioral strategies in novel or ambiguous situations such as when undertaking a memory task for the first time.[3] Thus, overall, the deficits are indicative of dysfunction in frontal–striatal circuit predominantly involving OFC, DLPFC, and caudate.

Clinical symptoms of OCD can be conceptualized in the background of these deficits; content of obsessions in OCD patients shares significant overlap with the content of intrusive thoughts experienced by healthy people. However, the difference between “normal” and “OCD” is that patients with OCD have failures to inhibit, or shift attention from these ongoing thoughts toward other more pleasant or less distressing cognitions. In addition, attention bias toward emotionally salient stimuli like “threat-related information” has been demonstrated in patients with OCD.[6225859] These findings are in support of the “threat-relatedness hypothesis” which proposes contents related to threat and associated cognitive processes to be the central characteristic of OCD.[60] Also, it has been suggested that treatment modifies meaning of these thoughts to that experienced by most other non-obsessional people.[61]

Results from different studies implicate failure in executive functions, memory (working and episodic), and laterality. These deficits have been shown to be sensitive to DLPFC, cingulate cortex, and parietal lobe. Importantly, same areas are reported to be abnormal in both structural and functional imaging studies. In addition, CC which plays an important role in interhemispheric transfer of information is also reported to be abnormal in patients with schizophrenia. Different models have been proposed to explain the clinical features of schizophrenia; most influential amongst them is the neurocognitive model proposed by Christopher Frith.[62] This model hypothesizes positive symptoms of schizophrenia to result from dysfunction of self-monitoring system and negative symptoms to result from defective spontaneous willed action. For example, first rank symptoms like somatic passivity might be secondary to parietal lobe abnormalities, whereas spontaneous willed action deficits might result from hypofrontality.[62]

Findings from neurocognitive and neuroimaging studies in bipolar disorder reveal impairments in hemispheric lateralization, executive functions, memory, and abnormal emotional perception. These functions are predominantly localized to prefrontal and limbic cortices. Importantly, affect regulation is considered to be the function of interacting prefrontal–limbic network with a top-down inhibition from the prefrontal cortex.[63] Imaging studies also report involvement of the same areas, namely, prefrontal cortex, amygdale, hippocampus, cerebellum, and basal ganglia. Thus, based on these findings, conceptual model for bipolar disorder considers prefrontal cortex and amygdale as the key components in affect regulation. According to frontal cortex–amygdala model, deficits in these and associated structures like cerebellum, hippocampus, and basal ganglia result in symptoms of bipolar disorder; hyperactivity of amygdala (bottom-up) or hypoactivity of prefrontal cortex (disrupted top-down) can result in affect dysregulation.[45]

THERAPEUTIC APPLICATIONS

One of the novel therapeutic options is the use of atomoxetine in the treatment of cognitive deficits of schizophrenia. Atomoxetine, which acts at dopaminergic and noradrenergic neurons of prefrontal cortex, is shown to improve attention and concentration in childhood disorders. We demonstrated the use of atomoxetine in treatment for cognitive deficits associated with schizophrenia.[64] Similarly, donepezil, a drug widely used in the treatment of Alzheimer's dementia, has been shown to be useful in treating and preventing cognitive deficits. Preliminary report from our group suggests donepezil as a potential therapeutic agent for treatment of cognitive deficits associated with electro convulsive therapy (ECT). Its prophylactic role in ECT-associated complications can have potential clinical utility.[65]

CONCLUSIONS AND FUTURE DIRECTIONS

Neurocognitive models provide a bridging link between brain pathology and phenomenology. They provide a useful framework to understand the pathogenesis of psychiatric disorders, bringing together isolated findings in behavioral, neuroimaging, and other neurobiological studies. They also have potential therapeutic application in formulating new pharmacological agents for clinical indications.